GBR-006

Hey Geraldine

Download PDF
United Kingdom Europe & Central Asia High income Pilot / Controlled Trial Phase Confirmed

Peterborough City Council

At a Glance

What it does LLMs for content creation, transformation and modality conversion — User communication and interaction
Who runs it Peterborough City Council
Programme Peterborough City Council Adult Social Care
Confidence Confirmed
Deployment Status Pilot / Controlled Trial Phase
Key Risks Model-related risks
Key Outcomes Over 1,200 conversations during 6-week trial.
Source Quality 4 sources — News article / media, Government website / press release

Hey Geraldine is a bespoke artificial intelligence assistant developed by Peterborough City Council in partnership with Outcomes Matter Consulting (AI consultant) and Datnexa (impact innovation technology partner), an AWS partner, Bloom accredited supplier, and member of the NatWest Accelerator programme. The system is designed to provide real-time, personalised information and guidance to occupational therapists and social workers in the council's adult social care service. It is named after Geraldine Jinks, a real therapy practitioner with 35 years of professional experience in the field, whose expertise the system aims to replicate and make available at scale across the workforce. Geraldine worked directly with the development team to feed information into the system so the AI could answer questions in her 'chatty and direct manner'.

The development of Hey Geraldine was motivated by a recognition that frontline adult social care staff frequently need rapid access to specialist knowledge about assistive technology and Technology Enabled Care (TEC) solutions when advising service users. Prior to the system's introduction, practitioners would need to consult colleagues, search through documentation, or rely on their own experience to answer open-ended practice queries such as what options are available for a person who forgets to turn off their oven, or what assistive technology might help someone with specific mobility challenges. Hey Geraldine was designed to handle precisely these kinds of unstructured, natural language queries by drawing on a curated knowledge base of domain-specific content about assistive technology and TEC solutions relevant to the Peterborough area.

The system was developed over a period of approximately three months, with a six-week trial period used to test and refine the system with practitioners. The project was funded through the Hospital Discharge fund, a government winter pressures funding mechanism. The council's deliberate decision to build a bespoke AI tool rather than adopt a commercial off-the-shelf solution was driven by concerns about data privacy and the need for the system to incorporate local nuance specific to Peterborough's adult social care landscape, which commercial products were judged unable to adequately address. The development process was characterised by close collaboration between council staff, Outcomes Matter Consulting, and Datnexa, with twice-weekly huddle meetings and structured feedback forms used to iteratively refine the system during the build phase.

Hey Geraldine employs sophisticated natural language processing capabilities to understand and respond to the varied and often vague queries that practitioners pose in their daily work. One of the key technical challenges encountered during development was balancing specificity with generalisation: the system needed to provide sufficiently targeted advice for individual cases while remaining applicable across the broad range of scenarios that adult social care professionals encounter. The AI analyses interactions to identify knowledge gaps across the workforce and provides an insights dashboard that tracks recurring themes, time savings, and areas where additional training or resources may be needed.

During its six-week trial, the system handled over 1,200 conversations with practitioners. Staff feedback was strongly positive, with users reporting that the system's responses were 'exactly as Geraldine would advise', indicating a high degree of fidelity to the expert knowledge it was designed to encode. Geraldine Jinks herself noted that 'some staff told me they thought they were conversing with me directly'. Quantitative evaluation found that the system saved approximately 15 minutes per conversation compared to traditional methods of seeking the same information, resulting in over 300 hours of total time saved during the trial period.

Prior to deployment, Peterborough City Council completed Data Protection Impact Assessments (DPIAs), Data Processing Agreements (DPAs), and service agreements to ensure the system met governance and compliance requirements. These formal assessments reflect the council's commitment to responsible AI deployment in a sensitive public service context where the information provided can directly affect the care and support available to vulnerable adults.

The system is currently in a prototype and testing phase, with plans to expand its availability to a broader set of users. The next phase of development is a self-serve version of Hey Geraldine for Peterborough residents, extending the system's reach beyond council practitioners to the general public. Additional future development plans include integration with Microsoft Teams to embed the assistant more seamlessly into practitioners' existing workflows, and the establishment of Technology Enabled Care champions within the council to promote adoption and provide peer support. The system is accessible via the HeyGeraldine.co.uk website. The Local Government Association has featured Hey Geraldine as a case study in its collection of innovative council AI implementations, recognising it as an example of how local authorities can develop tailored AI solutions to address specific service delivery challenges in adult social care.

Classifications follow the DCI AI Hub Taxonomy. Hover over field labels for definitions.

Social Protection Functions

Implementation/delivery chain
Case management primary
SP Pillar (Primary) The social protection branch: social assistance, social insurance, or labour market programmes. Social assistance
Programme Name Peterborough City Council Adult Social Care
Programme Type The type of social protection programme, classified under social assistance, social insurance, or labour market programmes. View in glossary Other
System Level Where in the social protection system the AI is applied: policy level, programme design, or implementation/delivery chain. View in glossary Implementation/delivery chain
Programme Description Peterborough City Council's adult social care service, which provides support to vulnerable adults including occupational therapy and Technology Enabled Care (TEC) assessments. Hey Geraldine is an AI assistant developed to support occupational therapists and social workers by providing real-time guidance on assistive technology and TEC solutions.
Implementation Type How the AI output is produced: Classical ML, Deep learning, Foundation model, or Hybrid. Affects validation, compute requirements, and governance profile. View in glossary Foundation model
Lifecycle Stage Current stage in the AI lifecycle, from problem identification through to monitoring, maintenance and decommissioning. View in glossary Integration and Deployment
Model Provenance Origin of the AI model: developed in-house, adapted from open-source, commercial/proprietary, or accessed via third-party API. View in glossary Commercial/proprietary
Compute Environment Where the AI system runs: on-premise, government cloud, commercial cloud, or edge/device. View in glossary Commercial cloud
Compute Provider The specific cloud or infrastructure provider hosting the AI system. AWS (inferred from Datnexa's AWS partner status)
Sovereignty Quadrant Classification of data and compute sovereignty: I (Sovereign), II (Federated/Hybrid), III (Cloud with safeguards), or IV (Shared Innovation Zone). View in glossary III — Compute-Intensive Cloud with safeguards
Data Residency Where the data used by the AI system is stored: domestic, regional, or international. View in glossary Not documented
Cross-Border Transfer Whether data crosses national borders, and if so, whether documented safeguards are in place. View in glossary Not documented
Decision Criticality The rights impact of the decision the AI supports. High criticality requires HITL oversight; moderate requires HOTL; low may operate HOOTL. View in glossary Low
Human Oversight Type Level of human involvement: Human-in-the-Loop (active review), Human-on-the-Loop (monitoring), or Human-out-of-the-Loop (periodic audit). View in glossary HOTL
Development Process Whether the AI system was developed fully in-house, through a mix of in-house and third-party, or fully by an external provider. View in glossary Mix of in-house and third-party
Highest Risk Category The most significant structural risk source identified: data, model, operational, governance, or market/sovereignty risks. View in glossary Model-related risks
Risk Assessment Status Whether a formal risk assessment, informal assessment, or independent audit has been conducted for this system. Formal assessment

Risk Dimensions

Market, sovereignty and industry structure risks
Operational and system integration risks

Impact Dimensions

Autonomy, human dignity and due process
  • DPIA/AIA conducted
  • Human oversight protocol
CategorySensitivityCross-System LinkageAvailabilityKey Constraints
Unstructured and text-based contentNon-personalSingle source (no linkage)Currently available and usedDomain-specific curated knowledge base covering assistive technology and Technology Enabled Care (TEC) solutions relevant to Peterborough's adult social care context

Datnexa (2024) 'Reflections on Developing an AI Assistant for Adult Social Care - "Hey Geraldine"', Datnexa Blog. Available at: https://www.datnexa.com/post/reflections-on-developing-an-ai-assistant-for-adult-social-care-hey-geraldine (Accessed: 27 March 2026).

View source News article / media

Datnexa (2024) 'The Story Behind Hey Geraldine's Development', Datnexa Blog. Available at: https://www.datnexa.com/post/the-story-behind-hey-geraldine-s-development (Accessed: 30 March 2026).

View source News article / media

Local Government Association (2024) 'Peterborough City Council: Hey Geraldine, a personalised AI assistant', LGA Case Studies, December. Available at: https://www.local.gov.uk/case-studies/peterborough-city-council-hey-geraldine-personalised-ai-assistant (Accessed: 27 March 2026).

View source Government website / press release

Peterborough City Council (2024) 'Council turns long-standing staff member into a Chatbot to support social workers', Peterborough City Council. Available at: https://www.peterborough.gov.uk/engagement-hub/council-turns-long-standing-staff-member-into-a-chatbot-to-support-social-workers (Accessed: 30 March 2026).

View source Government website / press release
Deployment Status How far the system has progressed into real-world operational use, from concept/exploration through to scaled and institutionalised. View in glossary Pilot / Controlled Trial Phase
Year Initiated The year the AI system was first initiated or development began. 2024
Scale / Coverage The scale and geographic or population coverage of the deployment. Prototype phase; over 1,200 conversations during 6-week trial by occupational therapists and social workers at Peterborough City Council; over 300 hours total time saved; next phase is self-serve version for Peterborough residents
Funding Source The source(s) of funding for the AI system development and deployment. Hospital Discharge fund (government winter pressures funding via Peterborough City Council)
Technical Partners External technology vendors, academic partners, or development partners involved. Outcomes Matter Consulting (AI consultant); Datnexa (impact innovation technology partner, AWS partner, Bloom accredited supplier, NatWest Accelerator member)
Outcomes / Results Over 1,200 conversations during 6-week trial. Staff reported that responses were 'exactly as Geraldine would advise'. Geraldine Jinks noted 'some staff told me they thought they were conversing with me directly'. Approximately 15 minutes saved per conversation, over 300 hours total time saved during the trial. Insights dashboard tracks themes, time savings, and knowledge gaps across the workforce.
Challenges Balancing specificity with generalisation: the system needed to provide targeted advice for individual cases while remaining applicable across a broad range of adult social care scenarios. Handling vague and open-ended queries from practitioners required sophisticated natural language processing. Commercial off-the-shelf solutions were rejected due to privacy concerns and inability to incorporate local nuance specific to Peterborough's adult social care landscape.

How to Cite

DCI AI Hub (2026). 'Hey Geraldine', AI Hub AI Tracker, case GBR-006. Digital Convergence Initiative. Available at: https://socialprotectionai.org/use-case/GBR-006 [Accessed: 1 April 2026].

Change History

Created 30 Mar 2026, 08:39
by v2-import (import)